Put the code for setting up and removing op blockers into an own
function, respectively. Then, we can invoke those functions whenever a
BDS is removed from an virtio-blk BB or inserted into it.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This reverts the changes that commit
2e1280e8ff applied to hw/block/fdc.c;
also, an additional case of drv->media_inserted use has crept in since,
which is replaced by a call to blk_is_inserted().
That commit changed tests/fdc-test.c, too, because after it, one less
TRAY_MOVED event would be emitted when executing 'change' on an empty
drive. However, now, no TRAY_MOVED events will be emitted at all, and
the tray_open status returned by query-block will always be false,
necessitating (different) changes to tests/fdc-test.c and iotest 118,
which is why this patch is not a pure revert of said commit.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-id: 1454096953-31773-4-git-send-email-mreitz@redhat.com
Reviewed-by: Eric Blake <eblake@redhat.com>
Background on QEMU boot indices
-------------------------------
Normally, the "bootindex" property is configured for bootable devices
with:
DEVICE_instance_init()
device_add_bootindex_property(..., "bootindex", ...)
object_property_add(..., device_get_bootindex,
device_set_bootindex, ...)
and when the bootindex is set on the QEMU command line, with
-device DEVICE,...,bootindex=N
the setter that was configured above is invoked:
device_set_bootindex()
/* parse boot index */
visit_type_int32()
/* verify unicity */
check_boot_index()
/* store parsed boot index */
...
/* insert device path to boot order */
add_boot_device_path()
In the last step, add_boot_device_path() ensures that an OpenFirmware
device path will show up in the "bootorder" fw_cfg file, at a position
corresponding to the device's boot index. Thus guest firmware (SeaBIOS and
OVMF) can try to boot off the device with the right priority.
NVMe boot index
---------------
In QEMU commit 33739c7129,
nvma: ide: add bootindex to qom property
the following generic setters / getters:
- device_set_bootindex()
- device_get_bootindex()
were open-coded for NVMe, under the names
- nvme_set_bootindex()
- nvme_get_bootindex()
Plus nvme_instance_init() was added to configure the "bootindex" property
manually, designating the open-coded getter & setter, rather than calling
device_add_bootindex_property().
Crucially, nvme_set_bootindex() avoided the final add_boot_device_path()
call. This fact is spelled out in the message of commit 33739c7129, and
it was presumably the entire reason for all of the code duplication.
Now, Vladislav filed an RFE for OVMF
<https://github.com/tianocore/edk2/issues/48>; OVMF should boot off NVMe
devices. It is simple to build edk2's existent NvmExpressDxe driver into
OVMF, but the boot order matching logic in OVMF can only handle NVMe if
the "bootorder" fw_cfg file includes such devices.
Therefore this patch converts the NVMe device model to
device_set_bootindex() all the way.
Device paths
------------
device_set_bootindex() accepts an optional parameter called "suffix". When
present, it is expected to take the form of an OpenFirmware device path
node, and it gets appended as last node to the otherwise auto-generated
OFW path.
For NVMe, the auto-generated part is
/pci@i0cf8/pci8086,5845@6[,1]
^ ^ ^ ^
| | PCI slot and (present when nonzero)
| | function of the NVMe controller, both hex
| "driver name" component, built from PCI vendor & device IDs
PCI root at system bus port, PIO
to which here we append the suffix
/namespace@1,0
^ ^
| big endian (MSB at lowest address) numeric interpretation
| of the 64-bit IEEE Extended Unique Identifier, aka EUI-64,
| hex
32-bit NVMe namespace identifier, aka NSID, hex
resulting in the OFW device path
/pci@i0cf8/pci8086,5845@6[,1]/namespace@1,0
The reason for including the NSID and the EUI-64 is that an NVMe device
can in theory produce several different namespaces (distinguished by
NSID). Additionally, each of those may (optionally) have an EUI-64 value.
For now, QEMU only provides namespace 1.
Furthermore, QEMU doesn't even represent the EUI-64 as a standalone field;
it is embedded (and left unused) inside the "NvmeIdNs.res30" array, at the
last eight bytes. (Which is fine, since EUI-64 can be left zero-filled if
unsupported by the device.)
Based on the above, we set the "unit address" part of the last
("namespace") node to fixed "1,0".
OVMF will then map the above OFW device path to the following UEFI device
path fragment, for boot order processing:
PciRoot(0x0)/Pci(0x6,0x1)/NVMe(0x1,00-00-00-00-00-00-00-00)
^ ^ ^ ^ ^ ^
| | | | | octets of the EUI-64 in address order
| | | | NSID
| | | NVMe namespace messaging device path node
| PCI slot and function
PCI root bridge
Cc: Keith Busch <keith.busch@intel.com> (supporter:nvme)
Cc: Kevin Wolf <kwolf@redhat.com> (supporter:Block layer core)
Cc: qemu-block@nongnu.org (open list:nvme)
Cc: Gonglei <arei.gonglei@huawei.com>
Cc: Vladislav Vovchenko <vladislav.vovchenko@sk.com>
Cc: Feng Tian <feng.tian@intel.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Cc: Kevin O'Connor <kevin@koconnor.net>
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Acked-by: Gonglei <arei.gonglei@huawei.com>
Acked-by: Keith Busch <keith.busch@intel.com>
Tested-by: Vladislav Vovchenko <vladislav.vovchenko@sk.com>
Message-id: 1453850483-27511-1-git-send-email-lersek@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
In Xen 4.7 we are refactoring parts libxenctrl into a number of
separate libraries which will provide backward and forward API and ABI
compatiblity.
One such library will be libxengnttab which provides access to grant
tables.
In preparation for this switch the compatibility layer in xen_common.h
(which support building with older versions of Xen) to use what will
be the new library API. This means that the gnttab shim will disappear
for versions of Xen which include libxengnttab.
To simplify things for the <= 4.0.0 support we wrap the int fd in a
malloc(sizeof int) such that the handle is always a pointer. This
leads to less typedef headaches and the need for
XC_HANDLER_INITIAL_VALUE etc for these interfaces.
Note that this patch does not add any support for actually using
libxengnttab, it just adjusts the existing shims.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Reviewed-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
The 2.88 drive is more suitable as a default because
it can still read 1.44 images correctly, but the reverse
is not true.
Since there exist virtio-win drivers that are shipped on
2.88 floppy images, this patch will allow VMs booted without
a floppy disk inserted to later insert a 2.88MB floppy and
have that work.
This patch has been tested with msdos, freedos, fedora,
windows 8 and windows 10 without issue: if problems do
arise for certain guests being unable to cope with 2.88MB
drives as the default, they are in the minority and can use
type=144 as needed (or insert a proper boot medium and omit
type=144/288 or use type=auto) to obtain different drive types.
As icing, the default will remain auto/144 for any pre-2.6
machine types, hopefully minimizing the impact of this change
in legacy hw to basically zero.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: John Snow <jsnow@redhat.com>
Message-id: 1453495865-9649-13-git-send-email-jsnow@redhat.com
This one is the crazy one.
fd_revalidate currently uses pick_geometry to tell if the diskette
geometry has changed upon an eject/insert event, but it won't allow us
to insert a 1.44MB diskette into a 2.88MB drive. This is inflexible.
The new algorithm applies a new heuristic to guessing disk geometries
that allows us to switch diskette types as long as the physical size
matches before falling back to the old heuristic.
The old one is roughly:
- If the size (sectors) and type matches, choose it.
- Fall back to the first geometry that matched our type.
The new one is:
- If the size (sectors) and type matches, choose it.
- If the size (sectors) and physical size match, choose it.
- Fall back to the first geometry that matched our type.
Signed-off-by: John Snow <jsnow@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 1453495865-9649-11-git-send-email-jsnow@redhat.com
2.88MB capable drives can accept 1.44MB floppies,
for instance. To rework the pick_geometry function,
we need to know if our current drive can even accept
the type of disks we're considering.
NB: This allows us to distinguish between all of the
"total sectors" collisions between 1.20MB and 1.44MB
diskette types, by using the physical drive size as a
differentiator.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: John Snow <jsnow@redhat.com>
Message-id: 1453495865-9649-10-git-send-email-jsnow@redhat.com
This patch adds a new explicit Floppy Drive Type option. The existing
behavior in QEMU is to automatically guess a drive type based on the
media inserted, or if a diskette is not present, arbitrarily assign one.
This behavior can be described as "auto." This patch adds the option
to pick an explicit behavior: 120, 144, 288 or none. The new "auto"
option is intended to mimic current behavior, while the other types
pick one explicitly.
Set the type given by the CLI during fd_init. If the type remains the
default (auto), we'll attempt to scan an inserted diskette if present
to determine a type. If auto is selected but no diskette is present,
we fall back to a predetermined default (currently 1.44MB to match
legacy QEMU behavior.)
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: John Snow <jsnow@redhat.com>
Message-id: 1453495865-9649-9-git-send-email-jsnow@redhat.com
Currently, QEMU chooses a drive type automatically based on the inserted
media. If there is no disk inserted, it chooses a 1.44MB drive type.
Change this behavior to be configurable, but leave it defaulted to 1.44.
This is not earnestly intended to be used by a user or a management
library, but rather exists so that pre-2.6 board types can configure it
to be a legacy value.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: John Snow <jsnow@redhat.com>
Message-id: 1453495865-9649-8-git-send-email-jsnow@redhat.com
Split apart pick_geometry by creating a pick_drive routine that will only
ever called during device bring-up instead of relying on pick_geometry to
be used in both cases.
With this change, the drive field is changed to be 'write once'. It is
not altered after the initialization routines exit.
media_validated does not need to be migrated. The target VM
will just revalidate the media on post_load anyway.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: John Snow <jsnow@redhat.com>
Message-id: 1453495865-9649-7-git-send-email-jsnow@redhat.com
pick_geometry is a convoluted function that makes it difficult to tell
at a glance what QEMU's current behavior for choosing a floppy drive
type is when it can't quite identify the diskette.
The code iterates over all entries in the candidate geometry table
("fd_formats") and if our specific drive type matches a row in the table,
then either "match" is set to that entry (an exact match) and the loop
exits, or "first_match" will be non-negative (the first such entry that
shares the same drive type), and the loop continues. If our specific
drive type is NONE, then all drive types in the candidate geometry table
are considered. After iteration, if "match" was not set, we fall back to
"first match".
This means that either "match" was set, or we exited the loop without an
exact match, in which case:
- If drive type is NONE, the default is truly fd_formats[0], a 1.44MB
type, because "first_match" will always get set to the first item.
- If drive type is not NONE, pick_geometry's iteration was fussier and
only looked at rows that matched our drive type. However, since all
possible drive types are represented in the table, we still know that
"first match" was set.
- If drive type is not NONE and the fd_formats table lists no options for
our drive type, we choose fd_formats[1], an incomprehensibly bizarre
choice that can never happen anyway.
Correct this: If first_match is -1, it can ONLY mean we didn't edit our
fd_formats table correctly. Throw an assertion instead.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: John Snow <jsnow@redhat.com>
Message-id: 1453495865-9649-6-git-send-email-jsnow@redhat.com
Currently, 'drive' is used both to represent the current diskette
type as well as the current drive type.
This patch adds a 'disk' field that is updated explicitly to match
the type of the disk.
As of this patch, disk and drive are always the same, but forthcoming
patches to change the behavior of pick_geometry will invalidate this
assumption.
disk does not need to be migrated because it is not user-visible state
nor is it currently used for any calculations. It is purely informative,
and will be rebuilt automatically via fd_revalidate on the new host.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: John Snow <jsnow@redhat.com>
Message-id: 1453495865-9649-5-git-send-email-jsnow@redhat.com
Change the floppy drive type to a QAPI enum type, to allow us to
specify the floppy drive type from the CLI in a forthcoming patch.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: John Snow <jsnow@redhat.com>
Message-id: 1453495865-9649-4-git-send-email-jsnow@redhat.com
Modify this function to operate directly on FDrive objects instead of
unpacking and passing all of those parameters manually. Reduces the
complexity in the caller and reduces the number of args to just one.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: John Snow <jsnow@redhat.com>
Message-id: 1453495865-9649-3-git-send-email-jsnow@redhat.com
Code motion: I want to refactor this function to work with FDrive
directly, so shuffle it below that definition.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: John Snow <jsnow@redhat.com>
Message-id: 1453495865-9649-2-git-send-email-jsnow@redhat.com
Move the ssi.h include file into the ssi directory.
While touching the code also fix the typdef lines as
checkpatch complains.
Signed-off-by: Alistair Francis <alistair.francis@xilinx.com>
Reviewed-by: Peter Crosthwaite <crosthwaite.peter@gmail.com>
Signed-off-by: Peter Crosthwaite <crosthwaite.peter@gmail.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Add the sst25wf080 SPI flash device.
Signed-off-by: Alistair Francis <alistair.francis@xilinx.com>
Reviewed-by: Peter Crosthwaite <crosthwaite.peter@gmail.com>
Signed-off-by: Peter Crosthwaite <crosthwaite.peter@gmail.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Clean up includes so that osdep.h is included first and headers
which it implies are not included manually.
This commit was created with scripts/clean-includes.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Done with this Coccinelle semantic patch
@@
expression FMT, E1, E2;
expression list ARGS;
@@
- error_setg(E1, FMT, ARGS, error_get_pretty(E2));
+ error_propagate(E1, E2);/*###*/
+ error_prepend(E1, FMT/*@@@*/, ARGS);
followed by manual cleanup, first because I can't figure out how to
make Coccinelle transform strings, and second to get rid of now
superfluous error_propagate().
We now use or propagate the original error whole instead of just its
message obtained with error_get_pretty(). This avoids suppressing its
hint (see commit 50b7b00), but I can't see how the errors touched in
this commit could come with hints. It also improves the message
printed with &error_abort when we screw up (see commit 1e9b65b).
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <1449764955-10741-3-git-send-email-armbru@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
It's necessary to distinguish source and target before we can add
blockdev-mirror, because we would want a concrete type of operation to
check on target bs before starting.
Signed-off-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 1450932306-13717-2-git-send-email-famz@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Request merging must not result in a huge request that exceeds the
maximum number of iovec elements. Use BlockLimits.max_iov instead of
hardcoding IOV_MAX.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
1. avoid possible superflous checking
2. make code more robustness
["make code more robustness" refers to avoiding integer
underflows/overflows.
--Stefan]
Signed-off-by: Gonglei <arei.gonglei@huawei.com>
Message-id: 1447207166-12612-1-git-send-email-arei.gonglei@huawei.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
src is stored in shared memory and src->nr_segments is dereferenced
twice at the end of the function. If a compiler decides to compile this
into two separate memory accesses then the size limitation could be
bypassed.
Fix it by removing the double access to src->nr_segments.
This is part of XSA-155.
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
The Xen toolstack uses "vhd" to specify a disk in VHD format, however
the name of the driver in QEMU is "vpc". Replace "vhd" with "vpc", so
that QEMU can find the right driver to use for it.
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
The official way of enabling dataplane is through the "iothread"
property that references an iothread object created by "-object
iothread". Since the old "x-data-plane=on" way now even crashes, it's
probably easier to just drop it:
$ qemu-system-x86_64 -drive file=null-co://,id=d0,if=none \
-device virtio-blk-pci,drive=d0,x-data-plane=on
ERROR:/home/fam/work/qemu/qom/object.c:1515:
object_get_canonical_path_component: assertion failed: (obj->parent != NULL)
Aborted
Signed-off-by: Fam Zheng <famz@redhat.com>
Message-id: 1449485967-19240-1-git-send-email-famz@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
For the "main area on file, oob in memory" case, fix the shifts so that
we erase the correct number of pages.
Signed-off-by: Ricard Wanderlöf <ricardw@axis.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This code has been dead for three years (since commit 7e7b7cba1).
Signed-off-by: Alberto Garcia <berto@igalia.com>
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Reviewed-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
"werror=report" would free the req in virtio_blk_handle_rw_error, we
mustn't write to it in that case.
Reported-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
Message-id: 1448239280-15025-1-git-send-email-famz@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
The shifts of the address mask and value shift beyond 32 bits when there
are 5 address cycles.
Cc: qemu-stable@nongnu.org
Signed-off-by: Rabin Vincent <rabin.vincent@axis.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Peter Crosthwaite <crosthwaite.peter@gmail.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
When a request R is absorbed by request M, it is appended to the
"mr_next" queue led by M, and is completed together with the completion
of M, in virtio_blk_rw_complete.
During DMA restart in virtio_blk_dma_restart_bh, requests in s->rq are
parsed and submitted again, possibly with a stale req->mr_next. It could
be a problem if the request merging in virtio_blk_handle_request hasn't
refreshed every mr_next pointer, in which case, virtio_blk_rw_complete
could walk through unexpected requests following the stale pointers.
Fix this by unsetting the pointer in virtio_blk_rw_complete. It is safe
because this req is either completed and freed right away, or it will be
restarted and parsed from scratch out of the vq later.
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Currently both BLKIF_OP_WRITE and BLKIF_OP_FLUSH_DISKCACHE are being
accounted as write operations.
Signed-off-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 7a2a14e3ac62027aa6267a6c02abc70717be9c0a.1446044837.git.berto@igalia.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
When creating snapshot with the dataplane enabled, the snapshot file gets
not the actual state of virtqueue, because the current state is stored in
VirtIOBlockDataPlane. Therefore, before saving snapshot need to sync
the dataplane vring state to the virtqueue. The dataplane will resume its
work at the next notify virtqueue.
When snapshot loads with loadvm we get a message:
VQ 0 size 0x80 Guest index 0x15f5 inconsistent with Host index 0x0:
delta 0x15f5
error while loading state for instance 0x0 of device
'0000:00:08.0/virtio-blk'
Error -1 while loading VM state
to reproduce the error I used the following hmp commands:
savevm snap1
loadvm snap1
qemu parameters:
--enable-kvm -smp 4 -m 1024 -drive file=/var/lib/libvirt/images/centos6.4.qcow2,if=none,id=drive-virtio-disk0,format=qcow2,cache=none,aio=native -device virtio-blk-pci,scsi=off,bus=pci.0,addr=0x8,drive=drive-virtio-disk0,id=virtio-disk0 -set device.virtio-disk0.x-data-plane=on
Signed-off-by: Pavel Butsykin <pbutsykin@virtuozzo.com>
Signed-off-by: Denis V. Lunev <den@openvz.org>
Message-id: 1445859777-2982-1-git-send-email-den@openvz.org
CC: Stefan Hajnoczi <stefanha@redhat.com>
CC: "Michael S. Tsirkin" <mst@redhat.com>
CC: Kevin Wolf <kwolf@redhat.com>
CC: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Devices that are compliant with virtio-1 do not support scsi
passthrough any more (and it has not been a recommended setup
anyway for quite some time). To avoid having to switch it off
explicitly in newer qemus that turn on virtio-1 by default, let's
switch the default to scsi=false for 2.5.
Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Message-id: 1444991154-79217-4-git-send-email-cornelia.huck@de.ibm.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
They will be excluded by type in the nested event loops in block layer,
so that unwanted events won't be processed there.
Signed-off-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
All callers pass in false, and the real external ones will switch to
true in coming patches.
Signed-off-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Jeff Cody <jcody@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
blk_bs() will not necessarily return a non-NULL value any more (unless
blk_is_available() is true or it can be assumed to otherwise, e.g.
because it is called immediately after a successful blk_new_with_bs() or
blk_new_open()).
Signed-off-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The tray of an FDD is open iff there is no medium inserted (there are
only two states for an FDD: "medium inserted" or "no medium inserted").
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Simplify memory allocation by sticking with a single API. GSlice
is not that fast anyway (tcmalloc/jemalloc are better).
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
The raw-posix block driver implements Linux AIO batching so multiple
requests can be submitted with a single io_submit(2) system call.
Batching is currently only used by virtio-scsi and
virtio-blk-data-plane.
Enable batching for regular virtio-blk so the number of io_submit(2)
system calls is reduced for workloads with queue depth > 1.
In 4KB random read performance tests with queue depth 32, the CPU
utilization on the host is reduced by 9.4%. The fio job is as follows:
[global]
bs=4k
ioengine=libaio
iodepth=32
direct=1
sync=0
time_based=1
runtime=30
clocksource=gettimeofday
ramp_time=5
[job1]
rw=randread
filename=/dev/vdb
size=4096M
write_bw_log=fio
write_iops_log=fio
write_lat_log=fio
log_avg_msec=1000
This benchmark was run on an raw image on LVM. The disk was an SSD
drive and -drive cache=none,aio=native was used.
Tested-by: Pradeep Surisetty <psuriset@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Symptom:
$ qemu-system-x86_64 -m 10000000
Unexpected error in ram_block_add() at /work/armbru/qemu/exec.c:1456:
upstream-qemu: cannot set up guest memory 'pc.ram': Cannot allocate memory
Aborted (core dumped)
Root cause: commit ef701d7 screwed up handling of out-of-memory
conditions. Before the commit, we report the error and exit(1), in
one place, ram_block_add(). The commit lifts the error handling up
the call chain some, to three places. Fine. Except it uses
&error_abort in these places, changing the behavior from exit(1) to
abort(), and thus undoing the work of commit 3922825 "exec: Don't
abort when we can't allocate guest memory".
The three places are:
* memory_region_init_ram()
Commit 4994653 (right after commit ef701d7) lifted the error
handling further, through memory_region_init_ram(), multiplying the
incorrect use of &error_abort. Later on, imitation of existing
(bad) code may have created more.
* memory_region_init_ram_ptr()
The &error_abort is still there.
* memory_region_init_rom_device()
Doesn't need fixing, because commit 33e0eb5 (soon after commit
ef701d7) lifted the error handling further, and in the process
changed it from &error_abort to passing it up the call chain.
Correct, because the callers are realize() methods.
Fix the error handling after memory_region_init_ram() with a
Coccinelle semantic patch:
@r@
expression mr, owner, name, size, err;
position p;
@@
memory_region_init_ram(mr, owner, name, size,
(
- &error_abort
+ &error_fatal
|
err@p
)
);
@script:python@
p << r.p;
@@
print "%s:%s:%s" % (p[0].file, p[0].line, p[0].column)
When the last argument is &error_abort, it gets replaced by
&error_fatal. This is the fix.
If the last argument is anything else, its position is reported. This
lets us check the fix is complete. Four positions get reported:
* ram_backend_memory_alloc()
Error is passed up the call chain, ultimately through
user_creatable_complete(). As far as I can tell, it's callers all
handle the error sanely.
* fsl_imx25_realize(), fsl_imx31_realize(), dp8393x_realize()
DeviceClass.realize() methods, errors handled sanely further up the
call chain.
We're good. Test case again behaves:
$ qemu-system-x86_64 -m 10000000
qemu-system-x86_64: cannot set up guest memory 'pc.ram': Cannot allocate memory
[Exit 1 ]
The next commits will repair the rest of commit ef701d7's damage.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <1441983105-26376-3-git-send-email-armbru@redhat.com>
Reviewed-by: Peter Crosthwaite <crosthwaite.peter@gmail.com>
* qemu_mutex_lock_iothread "No such process" fix
* cutils: qemu_strto* wrappers
* iohandler.c simplification
* Many other fixes and misc patches.
And some MTTCG work (with Emilio's fixes squashed):
* Signal-free TCG kick
* Removing spinlock in favor of QemuMutex
* User-mode emulation multi-threading fixes/docs
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
iQEcBAABCAAGBQJV8Tk7AAoJEL/70l94x66Ds3QH/3bi0RRR2NtKIXAQrGo5tfuD
NPMu1K5Hy+/26AC6mEVNRh4kh7dPH5E4NnDGbxet1+osvmpjxAjc2JrxEybhHD0j
fkpzqynuBN6cA2Gu5GUNoKzxxTmi2RrEYigWDZqCftRXBeO2Hsr1etxJh9UoZw5H
dgpU3j/n0Q8s08jUJ1o789knZI/ckwL4oXK4u2KhSC7ZTCWhJT7Qr7c0JmiKReaF
JEYAsKkQhICVKRVmC8NxML8U58O8maBjQ62UN6nQpVaQd0Yo/6cstFTZsRrHMHL3
7A2Tyg862cMvp+1DOX3Bk02yXA+nxnzLF8kUe0rYo6llqDBDStzqyn1j9R0qeqA=
=nB06
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/bonzini/tags/for-upstream' into staging
* Support for jemalloc
* qemu_mutex_lock_iothread "No such process" fix
* cutils: qemu_strto* wrappers
* iohandler.c simplification
* Many other fixes and misc patches.
And some MTTCG work (with Emilio's fixes squashed):
* Signal-free TCG kick
* Removing spinlock in favor of QemuMutex
* User-mode emulation multi-threading fixes/docs
# gpg: Signature made Thu 10 Sep 2015 09:03:07 BST using RSA key ID 78C7AE83
# gpg: Good signature from "Paolo Bonzini <bonzini@gnu.org>"
# gpg: aka "Paolo Bonzini <pbonzini@redhat.com>"
* remotes/bonzini/tags/for-upstream: (44 commits)
cutils: work around platform differences in strto{l,ul,ll,ull}
cpu-exec: fix lock hierarchy for user-mode emulation
exec: make mmap_lock/mmap_unlock globally available
tcg: comment on which functions have to be called with mmap_lock held
tcg: add memory barriers in page_find_alloc accesses
remove unused spinlock.
replace spinlock by QemuMutex.
cpus: remove tcg_halt_cond and tcg_cpu_thread globals
cpus: protect work list with work_mutex
scripts/dump-guest-memory.py: fix after RAMBlock change
configure: Add support for jemalloc
add macro file for coccinelle
configure: factor out adding disas configure
vhost-scsi: fix wrong vhost-scsi firmware path
checkpatch: remove tests that are not relevant outside the kernel
checkpatch: adapt some tests to QEMU
CODING_STYLE: update mixed declaration rules
qmp: Add example usage of strto*l() qemu wrapper
cutils: Add qemu_strtoull() wrapper
cutils: Add qemu_strtoll() wrapper
...
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
A number of files were including signal.h but not using any
of the functions it provides
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
